Geophys. Res. Lett., doi:10.1029/2011GL047908, 2011.

Ionospheric electron enhancement preceding the 2011 Tohoku-Oki earthquake

Kosuke Heki

Dept. Natural History Sci., Hokkaido University N10 W8, Kita-ku, Sapporo, Hokkaido 060-0810, Japan

Email: heki@mail.sci.hokudai.ac.jp

Abstract

The 2011 March 11 Tohoku-Oki earthquake (Mw9.0) caused vast damages to the country.

Large events beneath dense observation networks could bring breakthroughs to seismology

and geodynamics, and here I report one such finding. The Japanese dense network of Global

Positioning System (GPS) detected clear precursory positive anomaly of ionospheric total

electron content (TEC) around the focal region. It started ~40 minutes before the earthquake

and reached nearly ten percent of the background TEC. It lasted until atmospheric waves

arrived at the ionosphere. Similar preseismic TEC anomalies, with amplitudes dependent on

magnitudes, were seen in the 2010 Chile earthquake (Mw8.8), and possibly in the 2004

Sumatra-Andaman (Mw9.2) and the 1994 Hokkaido-Toho-Oki (Mw8.3) earthquakes, but not in

smaller earthquakes.

1. Introduction

Various kinds of earthquake precursors have been reported so far [Rikitake, 1976]. Among

others, electromagnetic phenomena have been explored worldwide, e.g. electric currents in the

ground [Uyeda and Kamogawa, 2008], propagation anomaly of VLF [Molchanov and

Hayakawa, 1998] and VHF [Moriya et al., 2010] radio waves. They are based on

measurements by purpose-built instruments at discrete observing points. Hence it has been

difficult to verify their spatial correlation with earthquakes. Spatial coverage could be

improved by satellite observations [Němec et al., 2008], but it is difficult to monitor the region

of future earthquakes continuously.

The 2011 March 11 (05:46UT) Tohoku-Oki earthquake ruptured the plate boundary ~450

km in length and ~200 km in width along the Japan Trench where the Pacific Plate subducts

beneath NE Japan (Fig.1). GEONET (GPS Earth Observation Network) is composed of >1000

continuous GPS stations. It has been yielding useful crustal deformation data since its launch

in 1994 [Heki, 2007], and already revealed coseismic and early postseismic crustal movements

of the Tohoku-Oki earthquake [Ozawa et al., 2011]. GEONET offers continuous

measurements of nearly two-dimensional crustal movements of the country. Although it could

detect mm level precursory crustal deformation of earthquakes, there have been no such

reports to date. GPS can also measure TEC, ionospheric electron contents integrated along

line-of-sights, by using the phase differences between the two L-band carrier waves. Here I

focus on the two-dimensional TEC distribution above Japan and its behavior immediately

before the 2011 Tohoku-Oki earthquake.

2. TEC Changes in the 2011 Tohoku Earthquake

A popular seismological target of GPS-TEC studies has been coseismic ionospheric

disturbances (CID) caused by various atmospheric waves generated by earthquakes [Calais

and Minster, 1995]. They include direct acoustic waves excited by vertical crustal movements

[Heki et al., 2006], Rayleigh surface waves [Rolland et al., 2011a], and internal gravity waves

[Occhipinti et al., 2008]. Fig.1A shows the TEC behavior before and after the earthquake seen

from a GPS station in NE Japan. At the time of earthquake (5:46UT), eight GPS satellites

were visible there (Fig.1B). CIDs are clearly seen with satellites 5, 15, 26, 27, 28 as the

irregular TEC changes caused by acoustic waves ~10 minutes after the earthquake, and with

satellites 18 and 22 as the oscillatory variations caused by internal gravity waves 40-80

minutes after the earthquake [Rolland et al., 2011b]. I try to isolate non-oscillatory TEC

anomalies immediately before the earthquake, which are not readily visible in Fig.1A.

Fig. 2 shows slant TEC changes over ~5 hours period observed at five GPS stations with

the satellite 15. The TEC shows gentle curvature due to satellite elevation changes. I employ

the data analysis procedure used for the detection of ballistic missile signatures in TEC [Ozeki

and Heki, 2010] (see Auxiliary Material). In addition to CID, transient positive anomaly of

TEC is seen to start ~40 minutes before the earthquake. Fig. 2B shows trajectories of

sub-ionospheric points (SIP) assuming a thin layer at 300 km altitude. The TEC anomaly is

larger for the SIP closer to the epicenter, and reaches ~4 TECU (1 TECU is 1016 electrons/m2).

On the other hand, slight negative TEC anomaly is seen at GPS stations with SIP distant from

the epicenter. These anomalies disappear and TEC returns to normal after CID arrivals.

Fig. 3 shows three snapshots of the geographical distribution of such TEC anomalies.

There are little anomalies 1 hour before the earthquake (Fig.3A). Positive anomalies have

already emerged in NE Japan 20 minutes before the earthquake (Fig.3B), and become larger

toward the earthquake occurrence time (Fig.3C). Slight negative changes of TEC are also seen

in SW Japan. The maximum positive anomaly of vertical TEC is ~2.3 TECU, which

corresponds to ~8 % of the background value (~27 TECU according to the global ionospheric

map [Mannucci et al., 1998] of this day). The latitudinal extent of the positive TEC anomaly

approximately overlaps with the ruptured fault segments. Similar figures to Figs. 2 and 3 were

drawn using three other satellites 27, 26, 9, and are shown in Figs. S1, 3, 5 and S2, 4, 6,

respectively. The spatial extent and amplitudes of the positive anomalies seems to vary

slightly from satellite to satellite. This would reflect the three-dimensional nature of the

electron density change which is projected onto a thin layer at 300 km altitude in drawing

these figures.

Global ionospheric maps (GIM) calculated from worldwide GPS stations have spatial

resolution of 2.5/5 degrees in latitude/longitude [Mannucci et al., 1998]. I downloaded the

model (codg0700.11i) from University of Berne (ftp.unibe.ch), and added the slant TEC

calculated for 0038/Sat.15 using this GIM to Fig.2A. Because the GPS station Mizusawa

(mizu in Fig.2B), only ~200 km from the epicenter, is used in deriving the GIM, the

synthesized curve also shows preseismic TEC enhancement although short wavelength

components are lost due to its low spatial resolution.

3. Discussion

3.1. Past earthquakes

In order to find out if such TEC anomalies commonly precede earthquakes, I analyzed

GPS data of the two recent M9 class earthquakes, the 2010 Central Chile (Maule) earthquake

(Mw8.8) [Moreno et al., 2010] and the Mw9.2 2004 Sumatra-Andaman earthquake [Banerjee et

al., 2005]. Raw data files of Argentine continuous GPS stations on Feb. 27, 2010, have been

downloaded from www.ign.gob.ar, and analyzed to see the TEC changes before the Chile

earthquake. The TEC time series obtained with the satellite 17, sensitive to ionosphere close to

the rupture area, show similar enhancement starting ~50 minutes before the earthquake and

recovery after the CID arrival (Fig. 4 and Fig.S9). The anomaly is 3-4 TECU in slant TEC,

somewhat smaller than the 2011 Tohoku earthquake.

CIDs of the 2004 Sumatra-Andaman earthquake have been analyzed using GPS stations in

Indonesia and Thailand [Heki et al., 2006]. SIP of the satellite 20 from Phuket (phkt) is very

close to the epicenter, and the slant TEC time series there is found to show temporary increase

exceeding 5 TECU lasting ~90 minutes (Figs.4 and S10). Although spatial coverage of GPS

data in these two earthquakes is not sufficient, preseismic positive TEC anomaly appears to

have accompanied these earthquakes.

In order to see if TEC enhancement occurs before smaller earthquakes, I analyzed

GPS-TEC data of several largest earthquakes in time and region covered by GEONET. In the

Mw8.3 1994 Hokkaido-Toho-Oki earthquake [Tsuji et al., 1995], the satellite 20 showed

similar positive TEC anomaly starting ~60 minutes before the earthquake (Figs.4 and S11).

Although significant, the amount of anomaly is much smaller than those of the three M9 class

earthquakes (Fig. 4). The 2003 Tokachi-Oki earthquake (Mw8.0) showed clear CID [Heki and

Ping, 2005]. This earthquake, together with a few other M7-8 class earthquakes with clear

CIDs, did not show significant preseismic TEC anomalies (Fig.S12). Fig.4B summarizes the

magnitude dependence of the vertical TEC enhancement immediately before the earthquake.

3.2. Models

There are no conclusive models for the preseismic electron enhancement. Here I give a few

clues for future searches of the model. The smaller anomalies for the satellites 9/27 than 15/26

(auxiliary material) would be explained if ionospheric F-layer electrons are assumed to have

moved down along the geomagnetic field. For example, we would see the positive TEC

anomaly similar to the observation if electrons within a layer of 20 km thickness at 350 km

height moved down to 300 km.

Electromagnetic earthquake precursors have been often attributed to positively charged

aerosols [Tributsch, 1978]. Several hypotheses have been proposed for their sources. For

example, experiments demonstrated that stresses mobilize positive holes in igneous rocks

[Takeuchi et al., 2006]. Alpha decay of radon released from the crust can also ionize the

atmosphere. They may change the electric resistivity of the lower atmosphere, which could

disturb the global electric circuit and redistribute ionospheric electrons [Pulinets and

Ouzounov, 2011]

The disappearance of the enhanced TEC after the passage of CID would be understood as

the mixing of ionosphere by acoustic waves. The amplitude of acoustic waves grows in the

light upper atmosphere, e.g. a few parts of 0.1 mm/s vertical velocity at the ground level is

amplified to >100 m/s at the F layer height [Rolland et al., 2011a]. Such high-speed

oscillations with periods of ~4.5 minute (atmospheric fundamental mode) would effectively

homogenize the electron density irregularities formed before earthquakes.

One may consider that the TEC decrease associated with CID caused apparent preseismic

TEC enhancement as an artifact. I demonstrate this is not the case in Figs. 4 and S3. I

extrapolated the TEC model fitted for the period before 5.2 UT (i.e. before the precursor

appeared) to the later period (gray dashed curves in Fig.4, and red curves in Fig.S3). The

observed TEC keeps deviating positively from the extrapolated curves during the ~40 minutes

period before the earthquake and CID brings them back to normal (i.e. it canceled out the

anomalous preseismic enhancement by stirring the F layer). This is also confirmed by

comparing the short-term TEC changing rates ~30 minutes before the earthquake between

regions near the epicenter and those farther away (Fig. S7). Fig. S5 also shows an example of

preseismic TEC enhancement without coseismic decrease.

4. Conclusions

Here I present an objectively testable scientific hypothesis that M9 class earthquakes are

immediately preceded by positive TEC anomalies of magnitude-dependent amplitude lasting

for an hour or so. Because the raw data files are available on the web, one can reproduce the

results reported here. It is also easy to apply the method to other earthquakes. The possible

precursor reported here is different qualitatively from past examples in three points, (1)

obvious temporal correlation (immediately before earthquakes), (2) obvious spatial correlation

(occurring around the focal area), and (3) clear magnitude dependence (Fig. 4B). The claim

that earthquakes are inherently unpredictable [e.g. Geller et al., 1997] might not be true at

least for M9 class earthquakes.

A few points need to be cleared before the preseismic TEC enhancement can be used for

short-term prediction of a large earthquake. A vital question is how often similar local TEC

enhancement of non-earthquake origins occur. In Fig. S8, I compare the five hours slant TEC

time series of the satellite 15 observed at 3009 from Jan.1 to Apr.30, 2011. They were all

modeled with degree-3 polynomials. The precursory anomaly of the Tohoku-Oki earthquake

was the largest of all. The second and the third largest anomalies (days 061 and 094) were

found to have traveled southward suggesting their space weather origin (Fig.S8E,F). In order

to discriminate such disturbances and preseismic TEC enhancements in real time, we may

need to monitor TEC outside the Japanese Islands.

A rather technical issue is the separation of the real TEC changes from apparent changes

due to satellite movements in the sky. At the moment, we need an arc of 1-2 hours or more to

accurately separate satellite-specific biases from temporal changes of TEC. The situation will

be improved by including Quasi-Zenith Satellite System (see qzss.jaxa.jp for detail). They are

supposed to stay near zenith and will enable real time observations of vertical TEC changes.

Acknowledgements The author is grateful to GSI, Japan, and IGN, Argentine, Manabu

Hashimoto (Kyoto Univ.) for providing the GPS data, the SEMS group members for

discussion. Reviews by three referees and the editor greatly improved the manuscript.

References

Banerjee, P., F. F. Pollitz, and R. Bürgmann (2005), The size and duration of the

Sumatra-Andaman earthquake from far-field static offsets, Science, 308, 1769-1772.

Calais, E., and. J. B. Minster (1995), GPS detection of ionospheric perturbations following the

January17, 1994, Northridge earthquake, Geophys. Res. Lett., 22, 1045-1048,

doi:10.1029/95GL00168.

Geller, R.J., D.D. Jackson, Y.Y. Kagan, and F. Mulargia (1997), Earthquakes cannot be

predicted, Science, 275, 1616-1617,.

Heki, K. (2007), Secular, transient and seasonal crustal movements in Japan from a dense GPS

array: Implication for plate dynamics in convergent boundaries, in The Seismogenic Zone

of Subduction Thrust Faults edited by T. Dixon and C. Moore, pp. 512-539, Columbia

University Press.

Heki, K. and J.-S. Ping (2005), Directivity and apparent velocity of the coseismic ionospheric

disturbances observed with a dense GPS array, Earth Planet. Sci. Lett., 236, 845-855.

Heki, K., Y. Otsuka, N. Choosakul, N. Hemmakorn, T. Komolmis, and T. Maruyama (2006),

Detection of ruptures of Andaman fault segments in the 2004 Great Sumatra Earthquake

with coseismic ionospheric disturbances, J. Geophys. Res., 111, B09313,

doi:10.1029/2005JB004202.

Mannucci, A.J., B.D. Wilson, D.N. Yuan, C.H. Ho, U.J. Lindqwister, and T.F. Runge (1998),

Radio Sci., 33, 565-582.

Molchanov, O.A. and M. Hayakawa (1998), VLF signal perturbations possibly related to

earthquakes, J. Geophys. Res., 103, 17489-17504.

Moreno, M., M. Bosenau, and O. Oncken (2010), 2010 Maule earthquake slip correlates with

pre-seismic locking of Andean subduction zone, Nature, 467, 198-202.

Moriya, T., T. Mogi, and M. Takada (2010), Anomalous pre-seismic transmission of

VHF-band radio waves resulting from large earthquakes, and its statistical relationship to

magnitude of impending earthquakes, Geophys. J. Int., 180, 858-870.

Němec, F., O. Santolík, M. Parrot, and J. J. Berthelier (2008), Spacecraft observations of

electromagnetic perturbations connected with seismic activity, Geophys. Res. Lett., 35,

L05109, doi:10.1029/2007GL032517.

Occhipinti, G., E. A. Kherani, and P. Lognonné (2008), Geomagnetic dependence of

ionospheric disturbances induced by tsunamigenic internal gravity waves, Geophys. J. Int.,

173, 753-765.

Ozawa, S., T. Nishimura, H. Suito, T. Kobayashi, M. Tobita, and T. Imakiire (2011),

Coseismic and postseismic slip of the 2011 magnitude-9 Tohoku-Oki earthquake, Nature,

474, doi:10.1038/nature10227.

Ozeki, M. and K. Heki (2010), Ionospheric holes made by ballistic missiles from North Korea

detected with a Japanese dense GPS array, J. Geophys. Res., 115, A09314,

doi:10.1029/2010JA015531.

Pulinets, S. and D. Ouzounov (2011), Lithosphere-atmosphere-ionosphere coupling (LAIC)

model – A unified concept for earthquake precursor validation, J. Asian Earth Sci. 41,

371-382, doi:10.1016/j.jseaes.2010.03.005.

Rikitake, T. (1976), Earthquake Prediction, Elsevier, Amsterdam.

Rolland, L. M., P. Lognonné, and H. Munekane (2011a), Detection and modeling of Rayleigh

wave induced patterns in ionosphere, J. Geophys. Res., 116, A05320,

doi:10.1029/2010JA016161.

Rolland, L. M., P. Lognonné, E. Astafyeva, E.A. Kherani, N. Kobayashi, M. Mann, and H.

Munekane (2011b), The resonant response of the ionosphere imaged after the 2011 off the

Pacific coast of Tohoku Earthquake, Earth Planets Space, 63, in press.

Takeuchi, A., B.W.S. Lau, and F. Freund (2006), Current and surface potential induced by

stress-activated positive holes in igneous rocks, Phys. Chem. Earth., 31, 240-247.

Tributsch, H. (1978), Do aerosol anomalies precede earthquakes? Nature, 276, 606-608.

Tsuji, H., Y. Hatanaka, T. Sagiya, and M. Hashimoto (1995), Coseismic crustal deformation

from the 1994 Hokkaido-Toho-Oki earthquake monitored by a nationwide continuous GPS

array in Japan, Geophys.Res. Lett., 22, 1669-1672.

Uyeda, S., M. Kamogawa (2008), The prediction of two large earthquakes in Greece, EOS.

Trans. Am. Geophys. Union, 89, 363.

Figure Captions

Figure 1. (A) Time series 3.5-8.5 UT of slant TEC changes observed at 0035 (black square in B)

with ten GPS satellites. Because each satellite-receiver pair has a different measurement bias, only

their temporal changes are meaningful. CIDs are seen in almost all satellites after the main shock at

5:46. (B) Trajectories of SIP for the satellites shown in (A). On the trajectories, small red stars are

SIP at 5:46 and small dots are hourly time marks. The rectangle shows the approximate rupture

zone of this earthquake. The large red star shows the epicenter.

Figure 2. (A) Slant TEC change time series taken at five GPS stations with the satellite 15.

Temporary positive TEC anomalies started ~40 minutes before the earthquake and disappeared

after CID passages. Black smooth curves are the models (see Auxiliary Material), and anomalies

are defined as the departure from the model curves. Slant TEC changes calculated using GIM for

site 0038 is shown as the blue curve. (B) Positions of the fives GPS stations (red dots) and their

5:00-6:00UT SIP trajectories (blue dots indicate 5:46). The northernmost point is 0038. Then

follow 0214, 3009, 0756 and 0596 southward.

Figure 3. Vertical TEC anomalies at three time epochs, (A) 1 hour, (B) 20 minutes, and (C) 1

minute before the earthquake, observed at GEONET stations with the satellite 15. Positive

anomalies (red color) are seen to grow near the focal region.

Figure 4. (A) Slant TEC changes and their models in the 2011 Tohoku earthquake, the 2004

Sumatra-Andaman earthquake, the 1994 Hokkaido-Toho-Oki earthquake, and the 2010 Chile

(Maule) earthquake. The horizontal axis shows the time from earthquakes. Dashed curves in gray

for the top two time series show the models derived with data prior to the possible onset of the

precursor (before 5.2 UT) and extrapolated to 5.2-6.0 UT. (B) Vertical TEC anomalies

immediately before the earthquakes as a function of their moment magnitudes. Colors correspond

to those in (A). In addition to the five data from four earthquakes in (A), three smaller earthquakes

(Fig. S12) are included (white circles).

Figure 1

Figure 2

Figure 3

Figure 4

Auxiliary Material

Text S1

Method:

A GPS station records L-band carrier phases in two frequencies, 1.5 (L1) and 1.2 (L2) GHz,

every 30 seconds (in GEONET). We downloaded the raw data on the day of the earthquake

from terras.gsi.go.jp. Temporal changes of the differences between the two phases expressed

in lengths are proportional to changes in TEC. To isolate the anomalous change of TEC, we model the raw TEC as a function of time t and angle ζ between line-of-sight and the local

zenith at its ionospheric penetration point with a model,

Slant TEC(t, ζ) = VTEC(t)/cosζ + e, (1)

where VTEC is the TEC when line-of-sight is perpendicular to the ionosphere. The bias e

inherent to phase observables of GPS remains constant for individual satellites in the studied period. VTEC can be separated from e with a few hours of observations because ζ varies with

time as the satellite moves in the sky (ζ can be calculated from orbital elements of GPS

satellites). VTEC changes diurnally and its change in a few hours period can be well

approximated with a cubic function of time t,

VTEC(t)=at3 + bt2 + ct + d. (2)

I estimated a,b,c,d and e using the least-squares method using the time series 3:00-8:00 UT.

In doing so, the period 5.2-6.0 UT, influenced by the preseismic TEC anomaly, was excluded

(similar exclusion was done for other earthquakes in Fig.4 and S12). The estimated models are

shown in Fig.2 with black smooth curves. The TEC anomaly was derived as the difference

between the model and the observed TEC, and converted to vertical TEC anomaly (Fig.3) by multiplying with cosζ.

Other satellites:

TEC changes in the 2011 Tohoku earthquake are plotted using the satellites 27 (Fig. S1),

26 (Fig. S3) and 9 (Fig. S5) in a similar manner to Fig. 2. The geographical distributions of

TEC anomalies at three time epochs are given in Fig. S2 (satellite 27), S4 (satellite 26), and S6

(satellite 9). It is interesting that the satellites 27 and 9 (Fig. S2, S6) show smaller TEC

anomalies than the satellites 15 (Fig.3) and 26 (Fig. S4). The angle between the line-of-sight

and the geomagnetic field is smaller for the satellites 27 and 9 (< 30 degrees) than the other

two satellites (50-70 degrees). Because ionospheric electron displacements in F layer are

largely constrained along the geomagnetic field, increased and decreased parts might have

cancelled each other considerably in integrating the electron content along the line-of-sight for

the satellites 27 and 9.

I demonstrate that the positive anomalies are not artifacts coming from coseismic TEC

decrease in three different ways. First, model curves inferred using the time series before 5.2

UT were extrapolated assuming linear changes of VTEC to later hours (Fig. S3). They indicate

that coseismic TEC decreases bring back the enhanced TEC to the state before the anomaly.

Secondly, five minutes average changing rates of vertical TEC a half hour before the

earthquake are mapped in Fig. S7. They show positive anomalies in NE Japan similar to

Fig.3C although they are not influenced by coseismic TEC decreases. Thirdly, the

northernmost station in Fig.S5 shows little coseismic TEC decrease (because its SIP at 05:45

UT is already away from the fault) although it shows clear preseismic positive anomaly.

Longer time window:

According to the space weather report (swc.nict.go.jp), geomagnetic activity on March 11

was relatively high due to the preceding corona hole activity. High geomagnetic activities

often give rise to wavy TEC disturbances generated in the auroral belt and traveling southward

(Large-scale traveling ionospheric disturbance, LSTID). However, local TEC enhancement

stationary above a certain area, like the preseismic TEC anomaly shown in Fig.3, cannot be

explained by such disturbances. It is also reported that strong sporadic E layer with the critical

frequency exceeding 8 MHz did not appear in Japan on that day.

Fig. 8A-D shows the five hours time series of slant TEC observed at 3009 (blue square in

Fig.8G) with the satellite 15 over the 4 months period in 2011. They were modeled using the

degree-3 polynomial without excluding any parts of the time series. The maximum deviation

from the model curve occurred on the day of the earthquake (red curve in Fig.8C, which is

identical to the middle curve in Fig.2). The second and the third largest deviations (days 094

and 061) are considered to be LSTID propagating southward across the whole Japanese

Islands. Fig. 8E, F shows their slant TEC time series at four GPS stations (red circles in

Fig.8G), and their propagation is recognized by time shifts of the positive anomaly peaks.

Other earthquakes:

TEC time series in the 2010 Chile (Maule) earthquake and the 2004 Sumatra-Andaman

earthquake are given in Figs. S9 and S10, respectively. In Fig.S10, the SIP of samp is closer to

the epicenter than phkt. However, the SIP of phkt is closer to the fault segment with the largest

moment release [Banerjee et al., 2005]. This would be the reason why phkt showed larger

preseismic TEC anomaly than samp. Fig.S11 shows the TEC time series in the 1994

Hokkaido-Toho-Oki earthquake observed with the satellite 20 at five GPS stations. The CID

of a few more earthquakes were observed with GEONET. They include the 2006 November

Kuril earthquake (Mw8.2) [Astafyeva and Heki, 2009], the 2003 Tokachi-Oki earthquake

(Mw8.0) and the foreshock of the 2004 September Off-Kii-Peninsula earthquake (Mw6.9)

[Heki and Ping, 2005]. As seen in Fig. S12, they show only small (1994 Hokkaido-Toho-Oki

and 2006 Kuril) or no (2003 Tokachi-Oki and 2004 Kii foreshock) preseismic TEC anomalies.

Auxiliary References:

Banerjee, P., F. F. Pollitz, and R. Bürgmann (2005), The size and duration of the

Sumatra-Andaman earthquake from far-field static offsets, Science, 308, 1769-1772.

Astafyeva, E. and K. Heki (2009), Dependence of waveform of near-field coseismic

ionospheric disturbance on focal mechanisms, Earth Planets Space, 61, 939-943.

Heki, K. and J.-S. Ping (2005), Directivity and apparent velocity of the coseismic ionospheric

disturbances observed with a dense GPS array, Earth Planet. Sci. Lett., 236, 845-855.

Heki, K. and K. Matsuo (2010), Coseismic gravity changes of the 2010 earthquake in Central

Chile from satellite gravimetry, Geophys. Res. Lett., 37, L24306,

doi:10.1029/2010GL045335.

Tsuji, H., Y. Hatanaka, T. Sagiya, and M. Hashimoto (1995), Coseismic crustal deformation

from the 1994 Hokkaido-Toho-Oki earthquake monitored by a nationwide continuous GPS

array in Japan, Geophys.Res. Lett., 22, 1669-1672.

Figure S1. (A) Slant TEC change time series observed at five GPS stations with the satellite

27 showing preseismic TEC enhancement similar to the satellite 15 (Fig.2). Black smooth

curves are the models derived assuming vertical TEC changing as cubic polynomials of time

(the time window used for the fit is 4.0-7.2 UT). The blue curve shows the slant TEC changes

calculated using GIM for site 0546. (B) Positions of the five GPS stations (red dots) in (A) and

their 5:00-6:00 SIP trajectories (blue dots show 5:46).

Figure S2. Snapshots of vertical TEC anomalies at (A) 1 hour, (B) 20 minutes, and (C) 1

minute before the earthquake observed with the satellite 27. Positive anomalies (red) are seen

in the region similar to Fig. 3. Note that the anomalies are somewhat smaller than the satellites

15 (Fig.3) and 26 (Fig. S4).

Figure S3. (A) Slant TEC change time series observed at five GPS stations with the satellite

26. Black smooth curves are the models derived assuming vertical TEC changing as cubic

polynomials of time (the time window used for the fit is 2.5-8.0 UT). The blue curve shows

the slant TEC changes calculated using GIM for site 0221. Dashed curves in red show the

models (polynomial degree 1) derived with data prior to the precursor onset (5.2 UT) and

extrapolated to the later period. (B) Positions of the five GPS stations (red dots) in (A) and their

5:00-6:00 SIP trajectories (blue dots show 5:46).

Figure S4. Vertical TEC anomalies at (A) 1 hour, (B) 20 minutes, and (C) 1 minute before the

earthquake observed with the satellite 26. Positive anomalies (red) are seen to extend partly to

the ocean.

Figure S5. (A) Slant TEC change time series observed at five GPS stations with the satellite 9.

Black smooth curves are the models derived assuming vertical TEC changing as cubic

polynomials of time (the time window used for the fit is 4.3-8.0 UT). The blue curve shows

the slant TEC changes calculated using GIM for site 0533. (B) Positions of the five GPS

stations (red dots) in (A) and their 5:00-6:00 SIP trajectories (blue dots show 5:46). The

station 0144 shows preseismic TEC enhancement but no coseismic decrease, suggesting that

the former is not an artifact caused by the latter.

Figure S6. Vertical TEC anomalies at (A) 1 hour, (B) 20 minutes, and (C) 1 minute before the

earthquake observed with the satellite 9. Note that the anomalies are smaller than the satellites

15 (Fig.3) and 26 (Fig. S4), and is similar to the satellite 27 (Fig. S2).

Figure S7. Average changing rate of vertical TEC over a five-minute period around 15:16 (30

minutes before the earthquake) of satellite 15 (derived from the time derivatives of slant TEC

curves in Fig. 2). Color shows the deviation of the rate from the average of all the stations. It is

seen that positive deviations are seen close to the rupture area.

Figure S8. Slant TEC time series (thick curves) over five hours period observed at 3009 (blue

square in G) with the satellite 15. A to D approximately correspond to January to April. Dst

(disturbance space-time) indices (average disturbance of the north component of geomagnetic

fields) are shown on the right-hand side. The time window is moved backwards two hours per

month because the GPS orbital period is a half sidereal day (i.e. appearance of the satellite 15

becomes earlier by ~4 minutes per day). Thinner curves show models where VTEC are

approximated with cubic functions of time (whole five hours periods shown in the figure are

used). A small arbitrary negative offset is given in plotting the model curves to avoid overlap

with the data. The root-mean-squares between the two curves were 0.34, 0.37, 0.56, 0.52

TECU for A, B, C, and D, respectively. The difference between the two curves exceeded 2.0

TECU (in VTEC) only once when preseismic TEC enhancement appeared (middle of the red

curve in C). The difference exceeded 1.5 TECU in two more occasions around 4 UT on the

day 061 and around 2 UT on the day 094 (marked with small black squares). They traveled

southward as recognized by comparing TEC time series at four points (red circles in G) in E

and F, respectively.

Figure S9. (A) Slant TEC change time series observed at six GPS stations in Chile/Argentine

with the satellite 17. Black smooth curves are the models derived assuming vertical TEC

changes as cubic polynomials of time (5.9-6.8 is excluded in estimating the model curves).

The 2010 February 27 Maule earthquake occurred at 6:34 UT. Its fault planes [Heki and

Matsuo, 2010] are shown by two rectangles in (B). Their 5.8-6.8 SIP trajectories (blue dots

indicate 6:34) are plotted as blue curves. Preseismic electron enhancement seems to have

started ~50 minutes before the earthquake as seen in the GPS stations, unsj, csj1, mzac, and

mzas. The TEC increase is smaller at sant and is not clear in lhcl.

Figure S10. (A) Slant TEC change time series observed at two GPS stations in Indonesia

(samp) and Thailand (phkt) with the satellite 20. Black smooth curves are the models derived assuming vertical TEC changing according to cubic polynomials of time (−0.5-1.2 UT is

excluded in estimating the model curves). The 2004 December 26 Sumatra-Andaman

earthquake occurred at 0:58 UT. Its fault planes [Banerjee et al., 2005] are shown by

rectangles in (B). Their 0.0-1.2 SIP trajectories (blue dots indicate 0:58) are plotted as blue

curves. Preseismic electron increase seems to have started ~90 minutes before the earthquake.

Figure S11. (A) Slant TEC change time series observed at five GPS stations in eastern Japan

with the satellite 20. Black smooth curves are the models derived assuming vertical TEC

changing according to cubic polynomials of time (12.5-13.8 UT is excluded in estimating the

model curves). The 1994 October 4 Hokkaido-Toho-Oki earthquake occurred at 13:22 UT. Its

fault plane [Tsuji et al., 1995] is shown in (B). Their 12.5-13.8 SIP trajectories (blue dots

indicate 13:22) are plotted as blue curves. Preseismic electron increase seems to have started

~50 minutes before the earthquake. TEC increase is much smaller than the three M9 class

earthquakes (Fig.4).

Figure S12. TEC behaviors within ±2 hours from the four M7-8 class earthquakes whose

CIDs have been studied with GEONET [Astafyeva and Heki, 2009; Heki and Ping, 2005].

Their SIPs are located within 200 km from the epicenter, and CIDs appear ~10 minutes after

the earthquakes. Weak TEC enhancements are seen before the 1994 (Mw8.3) and 2006

(Mw8.2) earthquakes, but not before the 2003 (Mw8.0) and 2004 (Mw6.9) earthquakes

(Fig.4B).